deg/sec looks just like radians/min TRAJ > ANGULAR_UNITS

More
29 Mar 2022 21:28 - 29 Mar 2022 21:30 #238664 by snakeoil
Config file attached. Work in progress but seems to work well.

4-axis router, XYZ + A, machine homes and jogs well etc... Except that angular units are inconsistent.

Under [TRAJ] section, MAX_ANGULAR_VELOCITY and DEFAULT_ANGULAR_VELOCITY have to be in RADIANS even though I've said "ANGULAR_UNITS = degree" (also tried "deg" as mentioned in INI docs)

Under [DISPLAY] (using gmoccapy) these have to be in DEGREES:
DEFAULT_ANGULAR_VELOCITY
MAX_ANGULAR_VELOCITY
MIN_ANGULAR_VELOCITY

Under [AXIS_A] again these have to be in RADIANS:
MAX_VELOCITY
MAX_ACCELERATION
[JOINT_4]
MAX_VELOCITY
MAX_ACCELERATION
STEPGEN_MAXVEL
STEPGEN_MAXACCEL

Version 2.8.2 from standard 32-bit Raspberry Pi image. All Gecko 201x and Nema 23 motors. Belt drive X/Y, ball screw Z, 3 home switches etc...

I am trying to debug another broader issue about speeds and G93/G94 but this seems more fundamental and maybe indicative of a larger problem.

Anyone else seeing this?
Attachments:
Last edit: 29 Mar 2022 21:30 by snakeoil.

Please Log in or Create an account to join the conversation.

More
29 Mar 2022 22:32 #238670 by BigJohnT
From the docs in the TRAJ section

ANGULAR_UNITS = <units> - Specifies the machine units for rotational axes. Possible choices are deg, degree (360 per circle), rad, radian (2pi per circle), grad, or gon (400 per circle). This does not affect the angular units of NC code. In RS274NGC, A-, B- and C- words are always expressed in degrees.


I'm not sure what you mean...

Please Log in or Create an account to join the conversation.

More
29 Mar 2022 22:33 #238671 by BigJohnT
MAX_VELOCITY
MAX_ACCELERATION
[JOINT_4]
MAX_VELOCITY
MAX_ACCELERATION
STEPGEN_MAXVEL
STEPGEN_MAXACCEL

The above is in units used per second.

JT

Please Log in or Create an account to join the conversation.

More
29 Mar 2022 23:43 #238680 by andypugh
I am pretty sure that LinuxCNC doesn't use radians. It is possible that it _looks_ like radians through a combination of the axis step scale and the difference between deg/min and deg/sec
The following user(s) said Thank You: snakeoil

Please Log in or Create an account to join the conversation.

More
30 Mar 2022 22:38 #238844 by snakeoil

I am pretty sure that LinuxCNC doesn't use radians. It is possible that it _looks_ like radians through a combination of the axis step scale and the difference between deg/min and deg/sec

I think this was a lot of my problem. 200 degrees per second is almost equal to 200 radians per minute.

Also there's a LOT of cruft in the INI files. Some of the params don't do anything. Others are not added by PNCConf.

TRAJ -> MAX_LINEAR_ACCELERATION is really important when doing simultaneous 3-axis milling on a rotary. It's not in the file but it's in the docs. This is, I guess, the sum of all moving axes. It defaults to 20.0 (!!!!)

Also "MAX_LINEAR_VELOCITY" under the DISPLAY section has no effect on gmoccapy. Gmoccay informs its UI based off of the value under the TRAJ section instead.

The entries under TRAJ for DEFAULT_ANGULAR_VELOCITY and MAX_ANGULAR_VELOCITY don't ever get referenced in the code, as far as I can tell, except this once in Axis.py:
default_jog_angular_speed = (
    inifile.find("DISPLAY", "DEFAULT_ANGULAR_VELOCITY")
    or inifile.find("TRAJ", "DEFAULT_ANGULAR_VELOCITY")
    or None)

Besides, if I think about it, what would "default" mean in terms of trajectory planning?

I really ought to take what I learned today and take the time to audit these params for 2.9

Please Log in or Create an account to join the conversation.

More
30 Mar 2022 23:27 #238848 by andypugh

I really ought to take what I learned today and take the time to audit these params for 2.9

If you have the time, that would be great. 
I assume that you know that HAL can reference the INI too (I don't think that this is relevant to anything that you have mentioned) 

Please Log in or Create an account to join the conversation.

Time to create page: 0.365 seconds
Powered by Kunena Forum